@loaders.gl/parquet 4.0.0-alpha.23 → 4.0.0-alpha.24

package/dist/lib/wasm/parse-parquet-wasm.js

@@ -1,27 +0,0 @@
-"use strict";
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.parseParquetWasm = void 0;
-const apache_arrow_1 = require("apache-arrow");
-const load_wasm_node_1 = require("./load-wasm/load-wasm-node");
-async function parseParquetWasm(arrayBuffer, options) {
-    const wasmUrl = options?.parquet?.wasmUrl;
-    const wasm = await (0, load_wasm_node_1.loadWasm)(wasmUrl);
-    const arr = new Uint8Array(arrayBuffer);
-    const arrowIPCUint8Arr = wasm.readParquet(arr);
-    const arrowIPCBuffer = arrowIPCUint8Arr.buffer.slice(arrowIPCUint8Arr.byteOffset, arrowIPCUint8Arr.byteLength + arrowIPCUint8Arr.byteOffset);
-    const arrowTable = tableFromIPC(arrowIPCBuffer);
-    return arrowTable;
-}
-exports.parseParquetWasm = parseParquetWasm;
-/**
- * Deserialize the IPC format into a {@link Table}. This function is a
- * convenience wrapper for {@link RecordBatchReader}. Opposite of {@link tableToIPC}.
- */
-function tableFromIPC(input) {
-    const reader = apache_arrow_1.RecordBatchStreamReader.from(input);
-    const recordBatches = [];
-    for (const recordBatch of reader) {
-        recordBatches.push(recordBatch);
-    }
-    return new apache_arrow_1.Table(recordBatches);
-}

package/dist/parquet-loader.js

@@ -1,41 +0,0 @@
-"use strict";
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.ParqueColumnnartLoader = exports.ParquetLoader = void 0;
-// __VERSION__ is injected by babel-plugin-version-inline
-// @ts-ignore TS2304: Cannot find name '__VERSION__'.
-const VERSION = typeof __VERSION__ !== 'undefined' ? __VERSION__ : 'latest';
-const DEFAULT_PARQUET_LOADER_OPTIONS = {
-    parquet: {
-        type: 'object-row-table',
-        url: undefined,
-        columnList: [],
-        geoparquet: true
-    }
-};
-/** ParquetJS table loader */
-exports.ParquetLoader = {
-    name: 'Apache Parquet',
-    id: 'parquet',
-    module: 'parquet',
-    version: VERSION,
-    worker: true,
-    category: 'table',
-    extensions: ['parquet'],
-    mimeTypes: ['application/octet-stream'],
-    binary: true,
-    tests: ['PAR1', 'PARE'],
-    options: DEFAULT_PARQUET_LOADER_OPTIONS
-};
-exports.ParqueColumnnartLoader = {
-    name: 'Apache Parquet',
-    id: 'parquet',
-    module: 'parquet',
-    version: VERSION,
-    worker: true,
-    category: 'table',
-    extensions: ['parquet'],
-    mimeTypes: ['application/octet-stream'],
-    binary: true,
-    tests: ['PAR1', 'PARE'],
-    options: DEFAULT_PARQUET_LOADER_OPTIONS
-};

package/dist/parquet-wasm-loader.js

@@ -1,26 +0,0 @@
-"use strict";
-// loaders.gl, MIT license
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.ParquetWasmLoader = void 0;
-// __VERSION__ is injected by babel-plugin-version-inline
-// @ts-ignore TS2304: Cannot find name '__VERSION__'.
-const VERSION = typeof __VERSION__ !== 'undefined' ? __VERSION__ : 'latest';
-/** Parquet WASM table loader */
-exports.ParquetWasmLoader = {
-    name: 'Apache Parquet',
-    id: 'parquet-wasm',
-    module: 'parquet',
-    version: VERSION,
-    worker: false,
-    category: 'table',
-    extensions: ['parquet'],
-    mimeTypes: ['application/octet-stream'],
-    binary: true,
-    tests: ['PAR1', 'PARE'],
-    options: {
-        parquet: {
-            type: 'arrow-table',
-            wasmUrl: 'https://unpkg.com/parquet-wasm@0.3.1/esm2/arrow1_bg.wasm'
-        }
-    }
-};

package/dist/parquet-wasm-writer.js

@@ -1,24 +0,0 @@
-"use strict";
-// loaders.gl, MIT license
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.ParquetWasmWriter = void 0;
-const encode_parquet_wasm_1 = require("./lib/wasm/encode-parquet-wasm");
-// __VERSION__ is injected by babel-plugin-version-inline
-// @ts-ignore TS2304: Cannot find name '__VERSION__'.
-const VERSION = typeof __VERSION__ !== 'undefined' ? __VERSION__ : 'latest';
-/** Parquet WASM writer */
-exports.ParquetWasmWriter = {
-    name: 'Apache Parquet',
-    id: 'parquet-wasm',
-    module: 'parquet',
-    version: VERSION,
-    extensions: ['parquet'],
-    mimeTypes: ['application/octet-stream'],
-    encode: encode_parquet_wasm_1.encode,
-    binary: true,
-    options: {
-        parquet: {
-            wasmUrl: 'https://unpkg.com/parquet-wasm@0.3.1/esm2/arrow1_bg.wasm'
-        }
-    }
-};

package/dist/parquet-writer.js

@@ -1,21 +0,0 @@
-"use strict";
-// loaders.gl, MIT license
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.ParquetWriter = void 0;
-// __VERSION__ is injected by babel-plugin-version-inline
-// @ts-ignore TS2304: Cannot find name '__VERSION__'.
-const VERSION = typeof __VERSION__ !== 'undefined' ? __VERSION__ : 'latest';
-exports.ParquetWriter = {
-    name: 'Apache Parquet',
-    id: 'parquet',
-    module: 'parquet',
-    version: VERSION,
-    extensions: ['parquet'],
-    mimeTypes: ['application/octet-stream'],
-    encodeSync,
-    binary: true,
-    options: {}
-};
-function encodeSync(data, options) {
-    return new ArrayBuffer(0);
-}

package/dist/parquetjs/codecs/declare.js

@@ -1,2 +0,0 @@
-"use strict";
-Object.defineProperty(exports, "__esModule", { value: true });

package/dist/parquetjs/codecs/dictionary.js

@@ -1,14 +0,0 @@
-"use strict";
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.encodeValues = exports.decodeValues = void 0;
-const rle_1 = require("./rle");
-function decodeValues(type, cursor, count, opts) {
-    opts.bitWidth = cursor.buffer.slice(cursor.offset, cursor.offset + 1).readInt8(0);
-    cursor.offset += 1;
-    return (0, rle_1.decodeValues)(type, cursor, count, { ...opts, disableEnvelope: true });
-}
-exports.decodeValues = decodeValues;
-function encodeValues(type, cursor, count, opts) {
-    throw new Error('Encode dictionary functionality is not supported');
-}
-exports.encodeValues = encodeValues;

package/dist/parquetjs/codecs/index.js

@@ -1,55 +0,0 @@
-"use strict";
-var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
-    if (k2 === undefined) k2 = k;
-    var desc = Object.getOwnPropertyDescriptor(m, k);
-    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
-      desc = { enumerable: true, get: function() { return m[k]; } };
-    }
-    Object.defineProperty(o, k2, desc);
-}) : (function(o, m, k, k2) {
-    if (k2 === undefined) k2 = k;
-    o[k2] = m[k];
-}));
-var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
-    Object.defineProperty(o, "default", { enumerable: true, value: v });
-}) : function(o, v) {
-    o["default"] = v;
-});
-var __importStar = (this && this.__importStar) || function (mod) {
-    if (mod && mod.__esModule) return mod;
-    var result = {};
-    if (mod != null) for (var k in mod) if (k !== "default" && Object.prototype.hasOwnProperty.call(mod, k)) __createBinding(result, mod, k);
-    __setModuleDefault(result, mod);
-    return result;
-};
-var __exportStar = (this && this.__exportStar) || function(m, exports) {
-    for (var p in m) if (p !== "default" && !Object.prototype.hasOwnProperty.call(exports, p)) __createBinding(exports, m, p);
-};
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.PARQUET_CODECS = void 0;
-const PLAIN = __importStar(require("./plain"));
-const RLE = __importStar(require("./rle"));
-const DICTIONARY = __importStar(require("./dictionary"));
-__exportStar(require("./declare"), exports);
-exports.PARQUET_CODECS = {
-    PLAIN: {
-        encodeValues: PLAIN.encodeValues,
-        decodeValues: PLAIN.decodeValues
-    },
-    RLE: {
-        encodeValues: RLE.encodeValues,
-        decodeValues: RLE.decodeValues
-    },
-    // Using the PLAIN_DICTIONARY enum value is deprecated in the Parquet 2.0 specification.
-    PLAIN_DICTIONARY: {
-        // @ts-ignore
-        encodeValues: DICTIONARY.encodeValues,
-        decodeValues: DICTIONARY.decodeValues
-    },
-    // Prefer using RLE_DICTIONARY in a data page and PLAIN in a dictionary page for Parquet 2.0+ files.
-    RLE_DICTIONARY: {
-        // @ts-ignore
-        encodeValues: DICTIONARY.encodeValues,
-        decodeValues: DICTIONARY.decodeValues
-    }
-};

package/dist/parquetjs/codecs/plain.js

@@ -1,211 +0,0 @@
-"use strict";
-var __importDefault = (this && this.__importDefault) || function (mod) {
-    return (mod && mod.__esModule) ? mod : { "default": mod };
-};
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.decodeValues = exports.encodeValues = void 0;
-const int53_1 = __importDefault(require("int53"));
-function encodeValues(type, values, opts) {
-    switch (type) {
-        case 'BOOLEAN':
-            return encodeValues_BOOLEAN(values);
-        case 'INT32':
-            return encodeValues_INT32(values);
-        case 'INT64':
-            return encodeValues_INT64(values);
-        case 'INT96':
-            return encodeValues_INT96(values);
-        case 'FLOAT':
-            return encodeValues_FLOAT(values);
-        case 'DOUBLE':
-            return encodeValues_DOUBLE(values);
-        case 'BYTE_ARRAY':
-            return encodeValues_BYTE_ARRAY(values);
-        case 'FIXED_LEN_BYTE_ARRAY':
-            return encodeValues_FIXED_LEN_BYTE_ARRAY(values, opts);
-        default:
-            throw new Error(`unsupported type: ${type}`);
-    }
-}
-exports.encodeValues = encodeValues;
-function decodeValues(type, cursor, count, opts) {
-    switch (type) {
-        case 'BOOLEAN':
-            return decodeValues_BOOLEAN(cursor, count);
-        case 'INT32':
-            return decodeValues_INT32(cursor, count);
-        case 'INT64':
-            return decodeValues_INT64(cursor, count);
-        case 'INT96':
-            return decodeValues_INT96(cursor, count);
-        case 'FLOAT':
-            return decodeValues_FLOAT(cursor, count);
-        case 'DOUBLE':
-            return decodeValues_DOUBLE(cursor, count);
-        case 'BYTE_ARRAY':
-            return decodeValues_BYTE_ARRAY(cursor, count);
-        case 'FIXED_LEN_BYTE_ARRAY':
-            return decodeValues_FIXED_LEN_BYTE_ARRAY(cursor, count, opts);
-        default:
-            throw new Error(`unsupported type: ${type}`);
-    }
-}
-exports.decodeValues = decodeValues;
-function encodeValues_BOOLEAN(values) {
-    const buf = Buffer.alloc(Math.ceil(values.length / 8));
-    buf.fill(0);
-    for (let i = 0; i < values.length; i++) {
-        if (values[i]) {
-            buf[Math.floor(i / 8)] |= 1 << i % 8;
-        }
-    }
-    return buf;
-}
-function decodeValues_BOOLEAN(cursor, count) {
-    const values = [];
-    for (let i = 0; i < count; i++) {
-        const b = cursor.buffer[cursor.offset + Math.floor(i / 8)];
-        values.push((b & (1 << i % 8)) > 0);
-    }
-    cursor.offset += Math.ceil(count / 8);
-    return values;
-}
-function encodeValues_INT32(values) {
-    const buf = Buffer.alloc(4 * values.length);
-    for (let i = 0; i < values.length; i++) {
-        buf.writeInt32LE(values[i], i * 4);
-    }
-    return buf;
-}
-function decodeValues_INT32(cursor, count) {
-    const values = [];
-    for (let i = 0; i < count; i++) {
-        values.push(cursor.buffer.readInt32LE(cursor.offset));
-        cursor.offset += 4;
-    }
-    return values;
-}
-function encodeValues_INT64(values) {
-    const buf = Buffer.alloc(8 * values.length);
-    for (let i = 0; i < values.length; i++) {
-        int53_1.default.writeInt64LE(values[i], buf, i * 8);
-    }
-    return buf;
-}
-function decodeValues_INT64(cursor, count) {
-    const values = [];
-    for (let i = 0; i < count; i++) {
-        values.push(int53_1.default.readInt64LE(cursor.buffer, cursor.offset));
-        cursor.offset += 8;
-    }
-    return values;
-}
-function encodeValues_INT96(values) {
-    const buf = Buffer.alloc(12 * values.length);
-    for (let i = 0; i < values.length; i++) {
-        if (values[i] >= 0) {
-            int53_1.default.writeInt64LE(values[i], buf, i * 12);
-            buf.writeUInt32LE(0, i * 12 + 8); // truncate to 64 actual precision
-        }
-        else {
-            int53_1.default.writeInt64LE(~-values[i] + 1, buf, i * 12);
-            buf.writeUInt32LE(0xffffffff, i * 12 + 8); // truncate to 64 actual precision
-        }
-    }
-    return buf;
-}
-function decodeValues_INT96(cursor, count) {
-    const values = [];
-    for (let i = 0; i < count; i++) {
-        const low = int53_1.default.readInt64LE(cursor.buffer, cursor.offset);
-        const high = cursor.buffer.readUInt32LE(cursor.offset + 8);
-        if (high === 0xffffffff) {
-            values.push(~-low + 1); // truncate to 64 actual precision
-        }
-        else {
-            values.push(low); // truncate to 64 actual precision
-        }
-        cursor.offset += 12;
-    }
-    return values;
-}
-function encodeValues_FLOAT(values) {
-    const buf = Buffer.alloc(4 * values.length);
-    for (let i = 0; i < values.length; i++) {
-        buf.writeFloatLE(values[i], i * 4);
-    }
-    return buf;
-}
-function decodeValues_FLOAT(cursor, count) {
-    const values = [];
-    for (let i = 0; i < count; i++) {
-        values.push(cursor.buffer.readFloatLE(cursor.offset));
-        cursor.offset += 4;
-    }
-    return values;
-}
-function encodeValues_DOUBLE(values) {
-    const buf = Buffer.alloc(8 * values.length);
-    for (let i = 0; i < values.length; i++) {
-        buf.writeDoubleLE(values[i], i * 8);
-    }
-    return buf;
-}
-function decodeValues_DOUBLE(cursor, count) {
-    const values = [];
-    for (let i = 0; i < count; i++) {
-        values.push(cursor.buffer.readDoubleLE(cursor.offset));
-        cursor.offset += 8;
-    }
-    return values;
-}
-function encodeValues_BYTE_ARRAY(values) {
-    // tslint:disable-next-line:variable-name
-    let buf_len = 0;
-    for (let i = 0; i < values.length; i++) {
-        values[i] = Buffer.from(values[i]);
-        buf_len += 4 + values[i].length;
-    }
-    const buf = Buffer.alloc(buf_len);
-    // tslint:disable-next-line:variable-name
-    let buf_pos = 0;
-    for (let i = 0; i < values.length; i++) {
-        buf.writeUInt32LE(values[i].length, buf_pos);
-        values[i].copy(buf, buf_pos + 4);
-        buf_pos += 4 + values[i].length;
-    }
-    return buf;
-}
-function decodeValues_BYTE_ARRAY(cursor, count) {
-    const values = [];
-    for (let i = 0; i < count; i++) {
-        const len = cursor.buffer.readUInt32LE(cursor.offset);
-        cursor.offset += 4;
-        values.push(cursor.buffer.slice(cursor.offset, cursor.offset + len));
-        cursor.offset += len;
-    }
-    return values;
-}
-function encodeValues_FIXED_LEN_BYTE_ARRAY(values, opts) {
-    if (!opts.typeLength) {
-        throw new Error('missing option: typeLength (required for FIXED_LEN_BYTE_ARRAY)');
-    }
-    for (let i = 0; i < values.length; i++) {
-        values[i] = Buffer.from(values[i]);
-        if (values[i].length !== opts.typeLength) {
-            throw new Error(`invalid value for FIXED_LEN_BYTE_ARRAY: ${values[i]}`);
-        }
-    }
-    return Buffer.concat(values);
-}
-function decodeValues_FIXED_LEN_BYTE_ARRAY(cursor, count, opts) {
-    const values = [];
-    if (!opts.typeLength) {
-        throw new Error('missing option: typeLength (required for FIXED_LEN_BYTE_ARRAY)');
-    }
-    for (let i = 0; i < count; i++) {
-        values.push(cursor.buffer.slice(cursor.offset, cursor.offset + opts.typeLength));
-        cursor.offset += opts.typeLength;
-    }
-    return values;
-}

package/dist/parquetjs/codecs/rle.js

@@ -1,145 +0,0 @@
-"use strict";
-// Forked from https://github.com/kbajalc/parquets under MIT license (Copyright (c) 2017 ironSource Ltd.)
-var __importDefault = (this && this.__importDefault) || function (mod) {
-    return (mod && mod.__esModule) ? mod : { "default": mod };
-};
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.decodeValues = exports.encodeValues = void 0;
-const varint_1 = __importDefault(require("varint"));
-// eslint-disable-next-line max-statements, complexity
-function encodeValues(type, values, opts) {
-    if (!('bitWidth' in opts)) {
-        throw new Error('bitWidth is required');
-    }
-    switch (type) {
-        case 'BOOLEAN':
-        case 'INT32':
-        case 'INT64':
-            // tslint:disable-next-line:no-parameter-reassignment
-            values = values.map((x) => parseInt(x, 10));
-            break;
-        default:
-            throw new Error(`unsupported type: ${type}`);
-    }
-    let buf = Buffer.alloc(0);
-    let run = [];
-    let repeats = 0;
-    for (let i = 0; i < values.length; i++) {
-        // If we are at the beginning of a run and the next value is same we start
-        // collecting repeated values
-        if (repeats === 0 && run.length % 8 === 0 && values[i] === values[i + 1]) {
-            // If we have any data in runs we need to encode them
-            if (run.length) {
-                buf = Buffer.concat([buf, encodeRunBitpacked(run, opts)]);
-                run = [];
-            }
-            repeats = 1;
-        }
-        else if (repeats > 0 && values[i] === values[i - 1]) {
-            repeats += 1;
-        }
-        else {
-            // If values changes we need to post any previous repeated values
-            if (repeats) {
-                buf = Buffer.concat([buf, encodeRunRepeated(values[i - 1], repeats, opts)]);
-                repeats = 0;
-            }
-            run.push(values[i]);
-        }
-    }
-    if (repeats) {
-        buf = Buffer.concat([buf, encodeRunRepeated(values[values.length - 1], repeats, opts)]);
-    }
-    else if (run.length) {
-        buf = Buffer.concat([buf, encodeRunBitpacked(run, opts)]);
-    }
-    if (opts.disableEnvelope) {
-        return buf;
-    }
-    const envelope = Buffer.alloc(buf.length + 4);
-    envelope.writeUInt32LE(buf.length, undefined);
-    buf.copy(envelope, 4);
-    return envelope;
-}
-exports.encodeValues = encodeValues;
-function decodeValues(type, cursor, count, opts) {
-    if (!('bitWidth' in opts)) {
-        throw new Error('bitWidth is required');
-    }
-    if (!opts.disableEnvelope) {
-        cursor.offset += 4;
-    }
-    let values = [];
-    while (values.length < count) {
-        const header = varint_1.default.decode(cursor.buffer, cursor.offset);
-        cursor.offset += varint_1.default.encodingLength(header);
-        if (header & 1) {
-            const count = (header >> 1) * 8;
-            values.push(...decodeRunBitpacked(cursor, count, opts));
-        }
-        else {
-            const count = header >> 1;
-            values.push(...decodeRunRepeated(cursor, count, opts));
-        }
-    }
-    values = values.slice(0, count);
-    if (values.length !== count) {
-        throw new Error('invalid RLE encoding');
-    }
-    return values;
-}
-exports.decodeValues = decodeValues;
-function decodeRunBitpacked(cursor, count, opts) {
-    // @ts-ignore
-    const bitWidth = opts.bitWidth;
-    if (count % 8 !== 0) {
-        throw new Error('must be a multiple of 8');
-    }
-    // tslint:disable-next-line:prefer-array-literal
-    const values = new Array(count).fill(0);
-    for (let b = 0; b < bitWidth * count; b++) {
-        if (cursor.buffer[cursor.offset + Math.floor(b / 8)] & (1 << b % 8)) {
-            values[Math.floor(b / bitWidth)] |= 1 << b % bitWidth;
-        }
-    }
-    cursor.offset += bitWidth * (count / 8);
-    return values;
-}
-function decodeRunRepeated(cursor, count, opts) {
-    // @ts-ignore
-    const bitWidth = opts.bitWidth;
-    let value = 0;
-    for (let i = 0; i < Math.ceil(bitWidth / 8); i++) {
-        // eslint-disable-next-line
-        value << 8; //  TODO - this looks wrong
-        value += cursor.buffer[cursor.offset];
-        cursor.offset += 1;
-    }
-    // tslint:disable-next-line:prefer-array-literal
-    return new Array(count).fill(value);
-}
-function encodeRunBitpacked(values, opts) {
-    // @ts-ignore
-    const bitWidth = opts.bitWidth;
-    for (let i = 0; i < values.length % 8; i++) {
-        values.push(0);
-    }
-    const buf = Buffer.alloc(Math.ceil(bitWidth * (values.length / 8)));
-    for (let b = 0; b < bitWidth * values.length; b++) {
-        if ((values[Math.floor(b / bitWidth)] & (1 << b % bitWidth)) > 0) {
-            buf[Math.floor(b / 8)] |= 1 << b % 8;
-        }
-    }
-    return Buffer.concat([Buffer.from(varint_1.default.encode(((values.length / 8) << 1) | 1)), buf]);
-}
-function encodeRunRepeated(value, count, opts) {
-    // @ts-ignore
-    const bitWidth = opts.bitWidth;
-    const buf = Buffer.alloc(Math.ceil(bitWidth / 8));
-    for (let i = 0; i < buf.length; i++) {
-        buf.writeUInt8(value & 0xff, i);
-        // eslint-disable-next-line
-        value >> 8; //  TODO - this looks wrong
-    }
-    return Buffer.concat([Buffer.from(varint_1.default.encode(count << 1)), buf]);
-}